Loading...

A Travelling Salesman Problem with Multi-Visit Drone

Saremi, Pantea | 2021

282 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54109 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Akbari Jokar, Mohammad Reza
  7. Abstract:
  8. Using drones in a synchronized last-mile delivery system with trucks has garnered attention in recent years. Due to various upsides of employing drones in delivery systems (namely, high speed and low environmental impact, etc.), it is considered a remarkable competitive advantage by many household brand in the E-commerce as well as urban logistics’ world. Furthermore, thanks to recent technological advances in producing higher capacity drones with longer flight range, this mode of delivery could now effectively exist alongside classic delivery means. A synchronized truck and drone problem, which was first introduced by Murray and Chu in 2015, has been exploited from various angles in the literature. The problem, which is originally called Flying Sidekick Traveling Salesman Problem (FSTSP), is defined as a truck with a drone on its roof that starts from a depot and serve a route consisting of customers. The drone could get dispatched from the truck on certain nodes, traverse a route of its own and serve a customer independently until it has to be retrieved by the truck again. This process is done as many times as needed until no more customers are left untended.In this study, a similar problem to the FSTSP is proposed titled “Travelling Salesman Problem with Multi-Visit Drone (TSP-MVD)” where drones are capable of serving more than one customer per each dispatch. This quality is one that has been vastly overlooked by most studies in the Literature of the synchronized truck and drone problem, despite the drone hardware advances making it fully possible. A mathematical MILP model has been developed to solve the problem in small instances, and two population-based metaheuristics (Genetic Algorithm (GA) and Dragonfly Algorithm (DA)) have been developed to tackle the problem for large instances. The ensuing results show consequential dominance of GA in comparison to DA for this specific problem. A sensitivity analysis has also been taken out on the two main parameters of the problem, further justifying the necessity of exploring this variation of the FSTSP problem
  9. Keywords:
  10. Genetic Algorithm ; Metaheuristic Method ; Traveling Salesman Problem ; Urban Logistics ; Dragonfly Algorithm ; Travelling Salesman Problem with Drone ; Truck and Drone Routing ; Last Mile Delivery

 Digital Object List

 Bookmark

No TOC